Perceptions of the Agency and Responsibility of the NHS COVID-19 App on Twitter: Critical Discourse Analysis.

BACKGROUND: Since September 2020, the National Health Service (NHS) COVID-19 contact-tracing app has been used to mitigate the spread of COVID-19 in the United Kingdom. Since its launch, this app has been a part of the discussion regarding the perceived social agency of decision-making algorithms. On the social media website Twitter, a plethora of views about the app have been found but only analyzed for sentiment and topic trajectories thus far, leaving the perceived social agency of the app underexplored. OBJECTIVE: We aimed to examine the discussion of social agency in social media public discourse regarding algorithm-operated decisions, particularly when the artificial intelligence agency responsible for specific information systems is not openly disclosed in an example such as the COVID-19 contact-tracing app. To do this, we analyzed the presentation of the NHS COVID-19 App on Twitter, focusing on the portrayal of social agency and the impact of its deployment on society. We also aimed to discover what the presentation of social agents communicates about the perceived responsibility of the app. METHODS: Using corpus linguistics and critical discourse analysis, underpinned by social actor representation, we used the link between grammatical and social agency and analyzed a corpus of 118,316 tweets from September 2020 to July 2021 to see whether the app was portrayed as a social actor. RESULTS: We found that active presentations of the app-seen mainly through personalization and agency metaphor-dominated the discourse. The app was presented as a social actor in 96% of the cases considered and grew in proportion to passive presentations over time. These active presentations showed the app to be a social actor in 5 main ways: informing, instructing, providing permission, disrupting, and functioning. We found a small number of occasions on which the app was presented passively through backgrounding and exclusion. CONCLUSIONS: Twitter users presented the NHS COVID-19 App as an active social actor with a clear sense of social agency. The study also revealed that Twitter users perceived the app as responsible for their welfare, particularly when it provided instructions or permission, and this perception remained consistent throughout the discourse, particularly during significant events. Overall, this study contributes to understanding how social agency is discussed in social media discourse related to algorithmic-operated decisions This research offers valuable insights into public perceptions of decision-making digital contact-tracing health care technologies and their perceptions on the web, which, even in a postpandemic world, may shed light on how the public might respond to forthcoming interventions.

The reception of public health messages during the COVID-19 pandemic.

Understanding the reception of public health messages in public-facing communications is of key importance to health agencies in managing crises, pandemics, and other health threats. Established public health communications strategies including self-efficacy messaging, fear appeals, and moralising messaging were all used during the Coronavirus pandemic. We explore the reception of public health messages to understand the efficacy of these established messaging strategies in the COVID-19 context. Taking a community-focussed approach, we combine a corpus linguistic analysis with methods of wider engagement, namely, a public survey and interactions with a Public Involvement Panel to analyse this type of real-world public health discourse. Our findings indicate that effective health messaging content provides manageable instructions, which inspire public confidence that following the guidance is worthwhile. Messaging that appeals to the audience's morals or fears in order to provide a rationale for compliance can be polarising and divisive, producing a strongly negative emotional response from the public and potentially undermining social cohesion. Provenance of the messaging alongside text-external political factors also have an influence on messaging uptake. In addition, our findings highlight key differences in messaging uptake by audience age, which demonstrates the importance of tailored communications and the need to seek public feedback to test the efficacy of messaging with the relevant demographics. Our study illustrates the value of corpus linguistics to public health agencies and health communications professionals, and we share our recommendations for improving the public health messaging both in the context of the ongoing pandemic and for future novel and re-emerging infectious disease outbreaks.

"The algorithm will screw you": Blame, social actors and the 2020 A Level results algorithm on Twitter.

In August 2020, the UK government and regulation body Ofqual replaced school examinations with automatically computed A Level grades in England and Wales. This algorithm factored in school attainment in each subject over the previous three years. Government officials initially stated that the algorithm was used to combat grade inflation. After public outcry, teacher assessment grades used instead. Views concerning who was to blame for this scandal were expressed on the social media website Twitter. While previous work used NLP-based opinion mining computational linguistic tools to analyse this discourse, shortcomings included accuracy issues, difficulties in interpretation and limited conclusions on who authors blamed. Thus, we chose to complement this research by analysing 18,239 tweets relating to the A Level algorithm using Corpus Linguistics (CL) and Critical Discourse Analysis (CDA), underpinned by social actor representation. We examined how blame was attributed to different entities who were presented as social actors or having social agency. Through analysing transitivity in this discourse, we found the algorithm itself, the UK government and Ofqual were all implicated as potentially responsible as social actors through active agency, agency metaphor possession and instances of passive constructions. According to our results, students were found to have limited blame through the same analysis. We discuss how this builds upon existing research where the algorithm is implicated and how such a wide range of constructions obscure blame. Methodologically, we demonstrated that CL and CDA complement existing NLP-based computational linguistic tools in researching the 2020 A Level algorithm; however, there is further scope for how these approaches can be used in an iterative manner.

Critical reflections on three popular computational linguistic approaches to examine Twitter discourses.

Although computational linguistic methods-such as topic modelling, sentiment analysis and emotion detection-can provide social media researchers with insights into online public discourses, it is not inherent as to how these methods should be used, with a lack of transparent instructions on how to apply them in a critical way. There is a growing body of work focusing on the strengths and shortcomings of these methods. Through applying best practices for using these methods within the literature, we focus on setting expectations, presenting trajectories, examining with context and critically reflecting on the diachronic Twitter discourse of two case studies: the longitudinal discourse of the NHS Covid-19 digital contact-tracing app and the snapshot discourse of the Ofqual A Level grade calculation algorithm, both related to the UK. We identified difficulties in interpretation and potential application in all three of the approaches. Other shortcomings, such the detection of negation and sarcasm, were also found. We discuss the need for further transparency of these methods for diachronic social media researchers, including the potential for combining these approaches with qualitative ones-such as corpus linguistics and critical discourse analysis-in a more formal framework.

Benchmarking framework for machine learning classification from fNIRS data.

Background: While efforts to establish best practices with functional near infrared spectroscopy (fNIRS) signal processing have been published, there are still no community standards for applying machine learning to fNIRS data. Moreover, the lack of open source benchmarks and standard expectations for reporting means that published works often claim high generalisation capabilities, but with poor practices or missing details in the paper. These issues make it hard to evaluate the performance of models when it comes to choosing them for brain-computer interfaces. Methods: We present an open-source benchmarking framework, BenchNIRS, to establish a best practice machine learning methodology to evaluate models applied to fNIRS data, using five open access datasets for brain-computer interface (BCI) applications. The BenchNIRS framework, using a robust methodology with nested cross-validation, enables researchers to optimise models and evaluate them without bias. The framework also enables us to produce useful metrics and figures to detail the performance of new models for comparison. To demonstrate the utility of the framework, we present a benchmarking of six baseline models [linear discriminant analysis (LDA), support-vector machine (SVM), k-nearest neighbours (kNN), artificial neural network (ANN), convolutional neural network (CNN), and long short-term memory (LSTM)] on the five datasets and investigate the influence of different factors on the classification performance, including: number of training examples and size of the time window of each fNIRS sample used for classification. We also present results with a sliding window as opposed to simple classification of epochs, and with a personalised approach (within subject data classification) as opposed to a generalised approach (unseen subject data classification). Results and discussion: Results show that the performance is typically lower than the scores often reported in literature, and without great differences between models, highlighting that predicting unseen data remains a difficult task. Our benchmarking framework provides future authors, who are achieving significant high classification scores, with a tool to demonstrate the advances in a comparable way. To complement our framework, we contribute a set of recommendations for methodology decisions and writing papers, when applying machine learning to fNIRS data.

Developing an Automated Assessment of In-session Patient Activation for Psychological Therapy: Codevelopment Approach.

BACKGROUND: Patient activation is defined as a patient's confidence and perceived ability to manage their own health. Patient activation has been a consistent predictor of long-term health and care costs, particularly for people with multiple long-term health conditions. However, there is currently no means of measuring patient activation from what is said in health care consultations. This may be particularly important for psychological therapy because most current methods for evaluating therapy content cannot be used routinely due to time and cost restraints. Natural language processing (NLP) has been used increasingly to classify and evaluate the contents of psychological therapy. This aims to make the routine, systematic evaluation of psychological therapy contents more accessible in terms of time and cost restraints. However, comparatively little attention has been paid to algorithmic trust and interpretability, with few studies in the field involving end users or stakeholders in algorithm development. OBJECTIVE: This study applied a responsible design to use NLP in the development of an artificial intelligence model to automate the ratings assigned by a psychological therapy process measure: the consultation interactions coding scheme (CICS). The CICS assesses the level of patient activation observable from turn-by-turn psychological therapy interactions. METHODS: With consent, 128 sessions of remotely delivered cognitive behavioral therapy from 53 participants experiencing multiple physical and mental health problems were anonymously transcribed and rated by trained human CICS coders. Using participatory methodology, a multidisciplinary team proposed candidate language features that they thought would discriminate between high and low patient activation. The team included service-user researchers, psychological therapists, applied linguists, digital research experts, artificial intelligence ethics researchers, and NLP researchers. Identified language features were extracted from the transcripts alongside demographic features, and machine learning was applied using k-nearest neighbors and bagged trees algorithms to assess whether in-session patient activation and interaction types could be accurately classified. RESULTS: The k-nearest neighbors classifier obtained 73% accuracy (82% precision and 80% recall) in a test data set. The bagged trees classifier obtained 81% accuracy for test data (87% precision and 75% recall) in differentiating between interactions rated high in patient activation and those rated low or neutral. CONCLUSIONS: Coproduced language features identified through a multidisciplinary collaboration can be used to discriminate among psychological therapy session contents based on patient activation among patients experiencing multiple long-term physical and mental health conditions.

The relationship between trust and attitudes towards the COVID-19 digital contact-tracing app in the UK.

During the COVID-19 pandemic, digital contact-tracing has been employed in many countries to monitor and manage the spread of the disease. However, to be effective such a system must be adopted by a substantial proportion of the population; therefore, public trust plays a key role. This paper examines the NHS COVID-19 smartphone app, the digital contact-tracing solution in the UK. A series of interviews were carried out prior to the app's release (n = 12) and a large scale survey examining attitudes towards the app (n = 1,001) was carried out after release. Extending previous work reporting high level attitudes towards the app, this paper shows that prevailing negative attitudes prior to release persisted, and affected the subsequent use of the app. They also show significant relationships between trust, app features, and the wider social and societal context. There is lower trust amongst non-users of the app and trust correlates to many other aspects of the app, a lack of trust could hinder adoption and effectiveness of digital contact-tracing. The design of technology requiring wide uptake, e.g., for public health, should embed considerations of the complexities of trust and the context in which the technology will be used.

A Review of Computer-Aided Expert Systems for Breast Cancer Diagnosis.

A computer-aided diagnosis (CAD) expert system is a powerful tool to efficiently assist a pathologist in achieving an early diagnosis of breast cancer. This process identifies the presence of cancer in breast tissue samples and the distinct type of cancer stages. In a standard CAD system, the main process involves image pre-processing, segmentation, feature extraction, feature selection, classification, and performance evaluation. In this review paper, we reviewed the existing state-of-the-art machine learning approaches applied at each stage involving conventional methods and deep learning methods, the comparisons within methods, and we provide technical details with advantages and disadvantages. The aims are to investigate the impact of CAD systems using histopathology images, investigate deep learning methods that outperform conventional methods, and provide a summary for future researchers to analyse and improve the existing techniques used. Lastly, we will discuss the research gaps of existing machine learning approaches for implementation and propose future direction guidelines for upcoming researchers.